DARPA has opened a Special Notice for a new brain-computer interface programme called Firefox. Program manager is Dr Pedro Irazoqui, based in DARPA’s Biological Technologies Office (BTO). Proposers Day is scheduled for 1 September 2026. Registration closes 21 August 2026.
Firefox targets a non-invasive optical brain-computer interface at 100 channels with 50-micron spatial resolution and 100-microsecond temporal resolution. Per DARPA’s verbatim programme objective as carried in The Sociable’s 17 August coverage: “Firefox aims to prove and de-risk the technology pathway to a 100-channel non-invasive brain-computer interface (BCI) with 50-micron spatial and 100-microsecond temporal resolution, replacing the need for highly invasive, surgically implanted electrodes in the BCIs of today.”
Firefox’s technical target
The scope is a coherent-optical measurement system that records neural signals at cortical depth in humans through an intact skull. The approach is distinct from EEG (which measures electrical activity at the scalp with limited spatial resolution) and from classical functional near-infrared spectroscopy (fNIRS), which measures haemodynamic proxies for neural activity on a slow timescale. Coherent-optical measurement uses the light itself, including phase and interferometric information, to reach neural tissue and pick up faster signals with finer spatial detail than either scalp-EEG or classical fNIRS can achieve non-invasively.
DARPA’s stated performance targets (100 channels at 50-micron spatial and 100-microsecond temporal resolution) are aimed at approaching the signal quality that surgically implanted intracortical electrodes deliver today, from outside the skull. If the targets are hit and replicated, the invasive-only assumption for high-bandwidth BCI weakens.
The two technical thrusts
Per DARPA’s programme description, Firefox has two named technical thrusts. The first is dynamic rastering and scanning to steer and expand the optical measurement across multiple regions of interest inside the brain rather than a fixed acquisition volume. The second is new data-processing architectures “beyond the digital domain,” reflecting the high data rate the optical instrument is expected to produce and the limits of standard digital pipelines at that rate. Photonic or analogue-domain processing paths are one plausible interpretation, but DARPA has not named a specific architectural approach in the public notice.
The N3 lineage
Firefox reads as the direct continuation of the coherent-optical track that DARPA first funded under its Next-Generation Nonsurgical Neurotechnology (N3) programme, publicly documented as announced in 2018 with six team awards issued in 2019. One of those teams, at the Johns Hopkins University Applied Physics Laboratory (JHU/APL) under Dr David Blodgett, publicly pursued coherent-optical measurement of neural activity as the JHU/APL N3 approach. Firefox’s language of “prove and de-risk the technology pathway” fits a six-year progression from N3’s exploratory bet to a Firefox-stage capability programme. The lineage is an inference from the modality overlap and the JHU/APL track record; DARPA has not published a formal “Firefox is the N3 follow-on” statement.
Where Firefox sits alongside SHINE
Firefox is not the same programme as SHINE. SHINE is a Defense Sciences Office (DSO) programme run by program manager Robin Bonomi, targeting non-invasive neuroplasticity for injury recovery across five modality domains (molecular engineering, neuromodulation, biomaterials, behavioural interventions, synthetic biology). Firefox is a BTO programme run by Pedro Irazoqui, targeting a coherent-optical BCI capability. The two run in different offices, under different program managers, targeting different modalities and scopes.
Two DARPA neurotech Special Notices inside two weeks (SHINE DARPA-SN-26-111 surfaced in InsideBCI coverage on 11 August, Firefox picked up in The Sociable on 17 August) mark a visible clustering of BTO and DSO neurotech activity. DARPA has historically produced the talent and technology pipeline that today’s US commercial BCI cohort was built on, from the HAPTIX, RE-NET and N3 era onward.
Undisclosed items
The award vehicle, budget envelope, programme duration, foreign-national eligibility rules, and expected number of performer teams are not disclosed in the sources available at time of writing. The solicitation number and full sam.gov contract-opportunity metadata are behind an authentication wall and could not be independently confirmed for this article. Firefox’s specific DARPA office assignment was not read from the sam.gov page directly; the BTO framing rests on Irazoqui’s own home-office listing on DARPA’s people page combined with the standard convention that a DARPA programme sits in its program manager’s home office.
What to watch
The Proposers Day briefing on 1 September 2026 will spell out the solicitation vehicle, the technical evaluation criteria, and the funding envelope. Awardee identity, once down-selected, will indicate whether JHU/APL retains the coherent-optical lead or whether other teams (potentially Kernel, Openwater, HRL Laboratories, or university-based photonics groups) have moved into contention.
Watch whether the two DARPA neurotech announcements this month grow into a sustained multi-office portfolio. A third programme announced from BTO, DSO or another DARPA office in the coming months would indicate a strategic push rather than two independent programme cycles happening to overlap.
Watch whether the “beyond-digital” data-processing thrust attracts photonic-computing or neuromorphic-processing performers who have not historically played in the DARPA neurotech ecosystem. New performer categories entering a DARPA capability programme are an early indicator that adjacent technology fields are being pulled into neurotechnology.